Trust
One of the more important types of decisions is deciding who to trust and about what, but I don’t think I’ve seen any discussion here about rational methods in that area.
Any suggestions?
One of the more important types of decisions is deciding who to trust and about what, but I don’t think I’ve seen any discussion here about rational methods in that area.
Any suggestions?
Use Outside Views more. What does their record in their profession look like? What do personality factors like Big Five predict?
I think there is a huge difference between the question “How much should I trust this person who is in my life already?” and “How can I find people likely to be trustworthy?” Big five seems like it would be useful for filtering people to add to your life, like if you were dating or running a job interview, whereas I think local (potentially mutable) details start to screen that abstraction out once the person is given and you’re trying to make much more particular decisions about the circumstances and nature of the relationship and mutual reliance within it.
Nancy clarified in another comment:
If someone is looking for a spouse (who in my mind would be your number one trusted human being on the planet in the long run) the big result I’m aware of is to avoid romantic entanglement with neurotic people. Marrying a neurotic person is one of a handful of things that defies happiness set point theory and causes a significant persistent decrease in a person’s happiness, and it also predicts divorce. Easily googleable pop science summary here.
Neurotics can be hard to rely on. They might be particularly depressed or may be angry at you during a week you really need them. They tend to have higher variance in performance tests, so sparkling performance at one time is less likely to translate to similarly high performance later. If those are factors you think of as related to “trust” then they are probably significant here.
What Big Five personality factors correlate with trustworthiness?
I know Conscientiousness inversely correlates with things like crime; I would be a little surprised if Agreeableness did not correlate either way; I suspect Openness probably inversely correlates with trustworthiness; and don’t know about Extraversion or Neuroticism.
This can be unpacked/dissolved.
First, I think of people/situation pairs rather than people. Specific situations influence things so much that one loses a lot by trying to think of people more abstractly; there is the danger of the fundamental attribution error.
Some people/situations are wrong more often than others are. Some people/situations lie more to others than others do. Some people/situations lie more to themselves than others do.
Some are more concerned with false positives, others with false negatives.
I also tend to think of people as components of decision making processes, as well as comprised of analogous decision making processes. Science takes advantage of this through the peer review process, which pits credulous humans against each other in attempts to prove each other’s ideas wrong, and it ultimately produces a body of knowledge each piece of which is unlikely to be false. It is the best input for anyone who instead cares about something slightly different, such as what is most likely to be true when false positives and false negatives would be similarly dangerous.
This is the source of my respect for Scott Adams (creator of Dilbert), which I’ve noticed is surprisingly prevalent if irregular among intelligent people I respect who have no particular reason to connect with anything having to do with office work or cubicles. It’s something that people either “get” or “don’t get,” like the orange joke. The man in an incomplete thinker, and many hundreds of millions of people are better decision makers than he, but as a member of a decision making group few could better come up with creative, topical, unique approaches to problems. Pair him with an intelligent, moderately critical mind and one would have a problem solving group better than one of two moderately intelligent and creative people.
Some people/situations produce more signal than others, others a better signal/noise ratio, some only advise when they are confident in their advice, some advise whenever they think it would have marginal gain, etc.
If you have an important decision to make, ask how to make the decision, not who should make it. Set up a person/situation network—even if the only person to trust is yourself (I have seen some research on patterns of decisions better made on a full bladder than an empty one, and vice versa. There is no you, there is only a you/situation (e.g. bladder) pair. Nothing corresponds to you/(no bladder situation, empty, full, or intermediate)! Likewise for decisions that differ dependent on whether or not your facial muscles are in the shape of a smile, etc.
Also, for every aspect of “trust,” beliefs are properly probabilistic; for the chances the person has good intentions, understands how you interpreted their words and actions, knows the right answer, knows they know the right answer, etc.
If you have a specific question you want advice to, asking about it most abstractly to avoid political associations was a great first move. Yet the abstract question is an imprecise summary and function of specific possible worlds. I think continuous rephrasing from more to less abstract might work well, as one could select from among variously abstract advice at different levels of political contamination and idiosyncratic specificity. Going in the other direction wouldn’t work as well, since the political content revealed early would taint later responses.
Here’s a useful heuristic: I don’t trust anyone who’s giving me advice that may hurt me and can’t possibly affect them. I’ll cross-check with another source, even if they’re an expert on the topic.
For instance, I didn’t trust my doctor when she gave me medication with a substantial chance of side-effects I wouldn’t have liked. I wouldn’t trust a real estate agent’s judgements on a house they’re trying to sell me. And so on.
That could be because your doctor was trying to prevent a nocebo effect. If a side effect is directly harmful (not simply a nuisance) or put you at risk say for example petic ulcer when taking large doses of NSAID:s e.g. aspirin, doctors are generally inclined to inform you—if they know their stuff.
Good, question. It depends some what by what you mean by trust and who. Do you mean strategies that apply across agent and/or person space or are things that rely on human specific psychology ok? Also, by trust do you mean reliably honest, reliable as an ally, or something else. I would be interested in all of those sub varieties, but which one is being addressed would be pretty important for any given article. Strategy of conflict is very good for how to ensure trust in a fairly unusual, but useful sense of the word.
I was thinking about which people you trust in important matters—spouses or the equivalent, employees, employers, doctors, investments, that sort of thing. If you’d like a LessWrongian flavor, who would you trust to take care of your affairs while you’re frozen, and why?
When it comes to doctors and therapists, my general approach is:
Seek recommendations from people I trust who are in a position to know. Try them out. If it’s not working, find someone else.
We’re on our 8th speech therapist.
Would you care to post about what can go wrong with speech therapists?
Do you mean material possessions or your body?
Either or both.
For material possessions, I plan to establish a trust and appoint a child or grandchild who is already signed up. Right now, I don’t trust either option with my body but will probably go with Alcor because of where I live.
Hmm. I think I know were to find some good social psyc data on related to that. I don’t think I can do it today, but I may be able to write a summary of it tomorrow.
The government security clearance manuals have documented what can be reduced to procedures and rules and whatnot. I know a guy who worked for the CIA a few years ago and he tells me the most trusted positions are the guys who do the security clearance evaluations. He said over half of them were Mormons. (Friend of a friend information is inherently untrustworthy.) One of the greatest spies in American history, James Angleton, was apparently paranoid to the brink of mental illness. It is generally a very difficult problem.
Upvoted for bringing up the intelligence community’s viewpoint on trust. I would say we could find some very interesting research on trust from that area. I think that because the intelligence community seems to be adversarial to a large degree. The problem of the double agent or mole, for example, would very likely lead intelligence agencies to invest heavily in metrics of trust.
The last job fair I went to I looked into a career with the CIA. I found they have extremely strict rules on who they hire, up to and including personality traits like patriotism.
You should find this econtalk podcast helpful. They are talking about David Rose’s book Moral Foundations of Economic Behavior and do a better job than most of defining what trust is and what it requires. And why it is important.
Thank you, though I wish there’d been more about what bright-line moral rules are a good enough deal for individuals that compliance makes sense, not to mention what can happen when a bright line system has good and bad parts, and people start to notice that.
As a result of going to that site, I listened to Nassim Taleb on anti-fragility (systems which become stronger when challenged)-- it’s at least as interesting.
Here is my general heuristic:
Whenever you have a question, find out what the consensus view is. Then see what the contrarians are saying. Then see what the consensus people are saying about what the contrarians are saying. See how the contrarians respond. Then make up your own mind.
I solved vaccines and cryonics this way.
By “mainstream consensus view”, I don’t mean what your average man on the street thinks. I mean what the experts (usually the “right” scientists) are saying on a topic. So creationism isn’t the consensus view. Evolution is.
Sometimes there isn’t really a consensus view. In that case start with what the contrarians are saying. Cryonics is like that.
Some people say they trust “hard sciences” but not “soft sciences”. But I think that isn’t right. When I’m trying to decide who to believe, I use something like the following, from most trustworthy to least:
Mathematicians, physicists, chemists, biologists, psychologists, climatologists, economists, anthropologists, historians, medical doctors, philosophers, people who write self-help books, people who write parenting books.
This is when I don’t have specific information about the person or group making the claim. The best philosopher is more reliable than the worst physicist.
I’m very interested as to why climatologists rank behind psychologists. I’m a little less surprised as to how low you rank medical doctors.
Edit: I’m not being snarky, I really would like to know the reasoning there.
When I wrote that I was thinking of people like Kahneman and Tversky. But you’re right. As a group, psychologists are less trustworthy.
People in general are not good epistemic peers. I think that is really dificult to encounter some caracteristic who define reliable people.
Intelligence is evidence of truth seeking. But if we are all cynical, the historic would help as a source of the type of intentions.
If you can get past the paywall, this article may be relevant.
http://www.tandfonline.com/doi/abs/10.1207/s15327795jra0903_1
Depends on how much information you have about someone, obviously, but one could look for situations with traditional game theory scenarios—may I’m just trying to use fancy words. When I’m in public places I sometimes watch people, trying to imaging (or just taking note) of what kind of person they are—just for fun, there is rarely any feedback but for example if a person goes through some trouble to but a book back on the right spot on a shelf in a library (in a situation where thinks it’s not being watched) I take that as quite good correlate for trustworthiness/ability to uphold social contracts.